Texts are one of the important elements that influence beginning reading instruction. In a recent study Shailaja Menon, University of Colorado/Boulder, and Elfrieda H. Hiebert, University of California/Berkeley, compared the effectiveness of a literature-based basal anthology text and a little-book curriculum in facilitating the word-reading skills of first graders.
Menon and Hiebert believe that different kinds of texts are essential for children’s development as readers and writers. Texts designed to scaffold word recognition skills could have different characteristics from those designed to support comprehension and love of reading. There has been little research that examines the effectiveness of different kinds of texts for beginning readers.
The features of a text that support beginning reading include pacing and repetition of words, sentence and text structures and decodable elements. Previous research has revealed two factors that affect beginning readers’ ability to decode text: cognitive load and linguistic content. Cognitive load refers to factors such as text predictability, word decodability and word density. Word density is the proportion of unique words in a text and the frequency of word repetition.
The length of the text influences cognitive load as well. Even small differences in text length and word density drastically change the nature of the task for beginning readers. Predictable syntactic and story patterns also reduce cognitive load for beginning readers. A close match between illustrations and the text reduces the cognitive load.
Linguistic content identifies critical word-level features that influence the difficulty level of the text. High-frequency or sight words and word patterns (rimes) all determine how difficult a passage will be for beginning readers. One hundred of the most common words in the English language account for more than half of the text in primary grades.
Beginning texts should provide systematic exposure to a core set of common phonograms or rimes (at, an, ill, et). Readers learn to generalize these when they see a variety of words with the same pattern. Decodability refers to how difficult words are for beginning readers to sound out.
Consonant-vowel and consonant-vowel-consonant are the two easiest patterns for beginning readers. Next easiest are short-vowel words with blends and digraphs, followed by long-vowel combinations. The least decodable are multi-syllabic and compound words that incorporate blends, digraphs and diphthongs. The pace at which new words are introduced, the number of times they are repeated and percentage of high-frequency words all influence the difficulty level of a text.
Four first-grade classrooms in one high-poverty, inner-city charter school participated in this study. Students were all African American. The school used a literature-based basal anthology — Houghton Mifflin’s Invitations to Literacy — that has been shown to be effective with similar student populations. Teachers in all four classrooms relied almost exclusively on this text during the first half of the school year. The two classrooms serving as comparisons continued to use this text for the 15 weeks of the study.
Two other classes used little books chosen from the Ready Reader program and sorted by these researchers for linguistic content and graded for cognitive load. Menon and Hiebert identified little books from the series by difficulty level. They were looking for books whose word and text characteristics would support students’ beginning reading in a systematic way. They were aware that this was something that teachers cannot be expected to do on their own.
Their goal was to identify differences between the little-book curriculum and the literature-based anthology and to determine if the restructured little-book curriculum was more effective in helping students acquire word-decoding skills. In addition, they studied whether the effects of these two reading series were different for children at different reading levels. Finally, they looked for differences in the percentage of children attaining grade-level reading by the end of the school year.
All children were tested before and after the study with word lists and graded reading passages from the Qualitative Reading Inventory. There were no significant differences in reading skills between classrooms at the beginning of the study. Students in each of the four classes were grouped into four reading groups on the basis of the pretest results.
Experimental and Comparison Texts
Researchers identified and sorted approximately 20 texts for linguistic content and cognitive load at each of the seven reading levels. Compared to the stories in the anthology text, these little books were much more predictable and sequenced in terms of difficulty. The texts of the little books showed a steady decrease in predictable features through first-grade levels, while the texts of the anthology vacillated. Some of the anthology stories were highly predictable even at the end of first grade. The little books consisted of 82% narratives and 18% expository texts. In contrast, the literature-based anthology text was entirely narrative.
Researchers observed each classroom three times weekly during the 15-week study. The observers noted the texts students read and other reading activities. When observers were not present, teachers kept logs of the texts read. In the research classrooms, students were assigned little books according to their assessed achievement level. Each group received five little books per week until all the books at that particular level were exhausted. Then they moved to the next higher level of the curriculum.
In comparison classrooms, all students read the same anthology text but teachers adjusted the pace of reading according the students’ achievement levels. Teachers ran their reading programs in their own ways and some classes covered more material than others. To account for potential differences in performance caused by the different amounts of total text read, post-test scores were analyzed by classroom.
All children in the experimental classrooms read more words and more different texts per week (except for the very lowest achievement group) than all students reading the anthology text. A higher percentage of words was repeated in the little books than in the anthology: between 85 and 90 percent of all words compared to 65 to 70 percent.
Greatest gains by high achievers
There was a significant difference in achievement between the experimental and regular classes. The groups reading little books improved by 2.8 reading levels, while the comparison classes improved only 1.8 reading levels. At the end of the 15-week intervention, the experimental classes were reading, on average, second-grade-level passages, while average comparison students were reading first-grade-level passages. The greatest gains were made by the high-achieving students; the average readers made the smallest gains.
By the end of the study, the very low- and low-achieving readers in the experimental classes were reading at similar levels to the average students in comparison classes. The average-reading students in the experimental classes were reading at the level of the high-achieving group in the comparison classes.
Analysis revealed that students reading the little-book curriculum made greater gains in reading words and texts than those reading the anthology, and there were more proficient readers who met first-grade benchmarks in the experimental classes. Two-thirds of the little-book students were reading above first-grade level at the end of the 15-week study.
Focus on target vowel patterns
The results of this study suggest that even a moderate amount of cognitive and linguistic support in texts can make a significant difference in the word-solving skills of first graders. Whether initially struggling, average or high readers, children in the little-book experiment read at one level higher than the students in the anthology groups after 15 weeks. Only 10 percent of little-book students had not attained grade level, while one-third of the students in the anthology-based reading classes were below grade level.
This suggests that somewhat greater consistency in linguistic content and less demanding cognitive loads support beginning reading acquisition. These results do not suggest that beginning readers require texts where all words fit particular patterns or where each unique word is repeated a particular number of times.
However, the significant difference between the anthology text and these little books was the percentage of unique words with common rimes and vowel patterns. A quarter of the words in the little books consistently exemplified target vowel patterns. Percentages were low in the anthology text. However, the percentage of words fitting common patterns in the little books was lower than those of earlier phonic book series.
The average repetition per word was twice as large for the experimental program: six repetitions for the little books and three for the anthology. Students in the little-book classrooms read more different words and stories per week than students in comparison classes. Anthology classes reread the same story several times throughout the week, reading about the same number of words, but not as many different stories as the little-book classes.
Both of the anthology teachers and one little-book teacher provided daily phonics and word-level instruction in addition to reading.
Demanding for educators
This study revealed that this first-grade anthology text does not consistently and gradually increase the linguistic complexity and difficulty of the reading text. These researchers acknowledge that literature-based anthologies may be valuable in improving the listening comprehension skills, vocabulary, and love of reading for first graders. Nevertheless, they do not attend systematically to the developmental needs of many first graders to acquire independent word-solving skills.
Curricular materials play a crucial role in shaping teachers’ instructional practices. The little-book curriculum used in this study required a substantial amount of time and effort by researchers and involved computer analysis of the word-level features of each text. Teachers cannot be expected to analyze the features of the texts their districts adopt. However, these little books were superior for improving both word-and passage-reading skills of poor, inner-city students in this study.
This study was limited to four classrooms in one school, and these researchers cannot show that these results will generalize to other settings. However, their study suggests that beginning reading texts that have been crafted to provide text- and word-level support can improve independent word solving and passage reading.
“A Comparison of First Graders’ Reading With Little Books or Literature-Based Basal Anthologies”, Reading Research Quarterly, Volume 40, Number 1, March 2005, pp. 12-38.
Published in ERN April 2005 Volume 18 Number 4